Collaborative Research: Improving Small Area Population Estimation with High-Resolution Remote Sensing

小区域人口估计对于理解和应对许多社会、政治、经济和环境问题至关重要。人口的规模和分布往往是州和地方政府资源分配的关键决定因素。在美国，通过全国人口普查，每十年只能获得一次详细的人口数据。这种频率往往不能满足快速增长地区的需求，比如德克萨斯州的奥斯汀，那里发生了值得注意的地方人口普查之间的变化。这个项目的总目标是发展一种成本效益高的方法，以产生详细的（普查区一级）、及时的（普查间）和准确的人口估计数。这将通过新兴的高空间分辨率遥感技术，如高空间分辨率卫星图像和光探测和测距（激光雷达），以自动方式改进住房单元估计来实现。然后将采用传统的住房单位（HU）方法来计算人口普查级别的人口数量。将进行两项单独的验证：包括家庭特别普查和与2010年人口普查的比较，以评估从开发的遥感方法得出的人口估计数的准确性。这个项目的目的是协同以前分别存在于遥感和人口地理学两个领域的努力，以便寻求解决办法，及时提供普查区一级的人口估计数。这项研究将对这两个领域的技术进步作出重大贡献，并证明在解决人口估计问题时将这两个领域联系起来的可行方案。拟议的方法将为人口预测提供更具成本效益的解决办法，因为该方法主要由遥感数据驱动，而不是由人口普查局辛苦收集的十年和调查数据驱动。此外，该项目的结果将与奥斯汀和附近快速发展地区的城市规划者和官员进行广泛讨论。

Small-area population estimates are essential for understanding and responding to many social, political, economic, and environmental problems. The size and distribution of the population often are key determinants for resource allocation for state and local governments. In the United States, detailed population data are only available for one date per decade through the national census. This frequency often does not meet the needs of rapid growth areas like Austin, Texas where noteworthy local inter-censal population changes occur. The overall objective of this project is to develop a cost-effective method to generate detailed (census block level), timely (inter-census) and accurate population estimates. This will be accomplished by development of an improved estimation of housing units in an automatic fashion from emerging high-spatial resolution remote sensing technologies, such as high spatial resolution satellite imagery and Light Detection And Ranging (LiDAR). A conventional Housing Unit (HU) method will then be employed to calculate population counts at census block level. Two separate validations: including a household special census, and comparisons to 2010 Census, will be performed to evaluate the accuracy of population estimates derived from the developed remote sensing methods. This project is intended to synergize previous efforts that separately existed in two fields: remote sensing and population geography, in order to seek solutions for providing timely population estimates at census block level. The research will make significant contributions to the technical advancement of the two fields as well as demonstrate a viable scheme for linking the two fields in addressing population estimation. The proposed method will provide a more cost-effective solution for population projections since the method is mainly driven by remote sensing data as compared to the laboriously collected decennial and survey data of the Census Bureau. In addition, the results of the project will be discussed extensively with city planners and officials from Austin and nearby rapidly growing localities.